Greg Palast has the article Florida = Honduras: Inequality kills. Want to end the American shooting epidemic?.
The result of my scatter plot came as quite the surprise to me: there was just about no correlation between number of guns and number of gun homicides.”
In fact, “the correlation coefficient was -0.105871699.” That is, by a small amount, more guns meant fewer homicides.
So what DID prove a strong correlation? Homicides versus the “GINI” coefficient. GINI is the measure of income inequality in a nation.
The graph he presents makes little sense as an explanation of the excerpt above. I have yet to figure out exactly how the horizontal axis of the graph should be labeled to make sense of it. We must also remember that correlation does not prove causation. However, at least in this case the premise does make sense to me. Given this hint of what these measures might be indicating, I’d like to see someone publish a real study of the possibilities.
Reluctantly, I have to file this story in the category of Greenberg’s Law Of The Media – If a news item has a number in it, then it is probably misleading.